tl;dr — Creating products in digital health is all about partnerships, the user ecosystem experience, and toeing a careful line with regulation. Interning as a PM at Watson Health has been an eye-opening experience and the company is more than meets the eye.
My internship: PM at Watson Health
This summer I interned as a Product Manager at Watson Health, a division in IBM that encompasses all of IBM’s healthcare products. I interned as part of IBM’s Extreme Blue program (it’s a bit different from the typical PM internship — feel free to read more about it here), meaning I worked on a nimble team with 3 developer interns in an incubator-like environment and was given nearly free-reign to ideate, prototype and demo new features for a cool existing Watson Health product for people with diabetes called Sugar.IQ (developed and distributed in partnership with Medtronic).
Throughout the summer, I interviewed users, conducted competitive research, and partnered with some seriously impressive developers, data scientists, and designers. We shipped a demo app that included a reimagined user experience, new features to make disease management easier ,and a new machine learning model to boot. We presented our final product to senior executives at IBM and Medtronic for consideration, with many aspects and concepts of our demo being adopted for the next release. Along the way, I was also able to talk to a PM on every single offering of Watson Health.
In the end, I had a few key takeaways on creating tech products in healthcare and Watson Health as a whole.
1. Partnership: It takes a village
Big tech companies interested in healthcare must have strong partners in the healthcare sector to help them better understand patient needs and navigate the complex incentives in the industry. Watson Health’s partnership with Medtronic helped us leverage their decades of experience and strong reputation in the diabetes space. This type of collaboration is critical and also comes with its additional complexity. Working in digital health means managing additional stakeholders and taking into account their strategic initiatives and technical infrastructure.
2. U(E)X: User ecosystem experience
Digital health products are not just for the user, but the user ecosystem. This summer our team had to consider how information in our product could be viewed easily by patients, caretakers, and clinicians. For example, people with diabetes may see their physician and dietitian once every 3 months. How can we make 90-day increments of information easily digestible for clinicians?
3. Regulation: A bulletproof system
In healthcare, sometimes it’s a matter of life and death. Unless products want to undergo the laborious FDA medical device approval process, they must be careful not to give medical advice that could contradict a clinician’s guidance or cause harm to a user. When testing a machine learning model, which can be hard to interpret, can we guarantee the output will be safe each time? If it’s not bulletproof, it’s not going in.
Watson Health: a force in digital health
There is no other big tech company with a commercialized product offering as diverse and comprehensive as Watson Health. Does Watson Health deserve its fair share of criticisms? Certainly. Many Watson Health products are in new and untested markets, resulting in a guaranteed amount of uncertainty. However, do I feel it has been singled out and portrayed by the media unfairly? Yes.
Aside: I’m often asked “what is Watson?”. How can one piece of technology do so much? Watson was formerly the machine that won Jeopardy, but since then it’s essentially become the umbrella name for all our cognitive products. It’s now a whole host of APIs and cognitive tools. Watson Health developers are not writing code for some monolithic “Watson”, but for new products that use AI / ML techniques.
A few things about Watson Health(WH) that I learned this summer:
- Huge and Growing: In 2017, WH reported 10,000 clients and partners. So far in 2018, it’s reporting 13,000 clients and partners and growing.
- More than Oncology: WH has products covering 1) Value-based care: integration of claims and clinical data to identify at-risk patients. 2) Imaging: enterprise image management and extraction / summarization of relevant patient information. 3) Government and public health : management of welfare / benefits enrollment and social workers’ case loads. 4) Genomics and life sciences: tools that match patients to clinical trials and aid researchers in new drug development.
- Evidence-based: More than 50 peer reviewed publications, posters, and abstracts support Watson Health cognitive offerings. The rigor and resources required to publish in academic journals is nothing to sneeze at, with many digital health products falling far short of this. Things aren’t perfect, but the company upholds itself to partnering tightly with clinicians and industry professionals to create its products.
Working under a microscope comes with working in healthcare at one of the largest and oldest tech companies around. People at Watson Health are truly passionate about their work and want to use data and technology to improve access, outcomes, and costs. They are willing to invest years and even decades of their lives (healthcare tech can have a long development and sales cycle) in pioneering aproduct that could help save lives.
Overall, this summer has been incredibly energizing. I’ve learned a ton and loved the experience. I can’t thank Watson enough.